自主无线电单元：Au-RUs.pdf

1、Brad DeforgeDell TechnologiesTelecoms Systems BusinessAutonomous Radio Units:Au-RUsLowest(CapEx)Cost at All Costs RUs(LCaAC)-Limited deployment scenarios-No innovation without profitabilityBenefits of expanded OpEx valuation-Energy Saving Features as a 1ststep-Performance optimization-Network level

2、deployment monitoring-Operator specific New datasets with high value data.Goal Full autonomy for lowest cost.Lowest Cost An Innovation Dead-EndEconomies of Lowest Cost at All CostsCommodity RULowest-Cost-at-All-Cost RUAdditional BOM/sBOMEnergy Saving Feature-Enabled RUDecreased OpExTCO for commodity

3、RUTCO for ESFRUAutonomous RUTCO for autonomousRUAdditional BOM/sBOMAutonomy:“Independence,freedom from external control or influence”.The promise:Generate and leverage data to lower overall costs through added value Mechanism for manufacturing automation in-field operational performance dataDeliver

4、RU operational,performance,and network dataRU performance,Network specific,and Operator specific dataML-driven strategies for lowest Total Cost of OwnershipOptimization strategies:Energy,ease of use,and maintenanceValue-added strategies:Channel sensing and modeling,user/asset tracking,and spectrum u

5、sageCustomer specific strategies:Use cases prediction algorithms or enhanced realityDevelop ML models not for AI factories,but for edge and N/W inferencingCost reduction:Dynamic tuning of parameters for optimizationCost avoidance:Self-healing and predictive methods for fault detection and resolution

6、Value-added functionality:Automated sensing,user tracking,and spectrum allocation management The Promise of AutonomyFocus:RU with ML inferencing capabilityGenerates data for targeted usesAggregate dataTrain ML modelsApply ML models 10 mS to 1 Sec35 uS to 10 mSMinutesRU with ML InferencingAutonomous